We propose to study the problem of the genetical changes that occur in species in which random extinction of local populations and the colonization of vacant habitats occur on the same time scale as the genetical changes within a single population. We will combine the results of a Monte Carlo simulation model with experimental results from laboratory populations of Tribolium to determine the effect of colonization and extinction on the possible rates and extent of genetic differentiation of local populations and to determine the conditions under which group selection could be an effective mechanism in natural populations. The model will be tested by comparison with several experimental results and the experiments will then provide estimates of the several parameters of the model. The model will then be applied to other systems to test the validity of our predictions. Our goal will be the development of a theory which is sufficiently general that it can be applied to natural systems but also sufficiently realistic that the underlying assumptions can be tested independently of the predictions.